A major problem in chemotherapy of cancer in man is the variation in response of a single histological type malignancy to a given drug. It is also a well-established fact that individual advanced transplanted murine tumors of similar size and histology exhibit responses to a drug that range from minimal to complete regression and cure. It is likely that the cytotoxic action of a drug (in the absence of cell resistance to the specific drug) is a function of the concentration times the duration of exposure. We propose to investigate the tumor and host factors that determine drug distribution within the tumor and clearance from the tumor. Direct correlation of tumor response with drug activation and excretion, drug levels in plasma and tumor in addition to tumor vascularity and proliferation at the time of initial treatment is not feasible because these tests require sacrifice of the host prior to tumor response. For initial experiments, we propose to treat tumorbearing mice with the first dose of drug on a protocol that requires multiple doses to obtain optimum response but where initial response is evident after the first treatment. From these treated mice we will select those that exhibit maximum and minimum response to use for evaluation of drug distribution in hosts and tumors at the time of second treatment. From these studies, we expect to determine whether the major variation may be attributed to host or tumor and to gain sufficient information to develop a noninvasive test to select potentially responsive and nonresponsive tumors from a group of untreated staged transplanted tumors. This predictive test would make feasible studies of tumors and host parameters that determine tumor response to initial treatment with a given drug. This information will be used to guide modifications of drug doses, schedules, and routes of administration for the nonresponsive tumors.